2g 3g phone ntworks

7/30/2019 2G 3G Phone Ntworks

1/20

Life cycle assessment of second generation (2G) and third generation (3G)mobile phone networks

Wolfram Scharnhorsta,b,, Lorenz M. Hilty b, Olivier Jollieta,c

a Life Cycle Systems Group (ENAC-GECOS)/Swiss Federal Institute of Technology, CH-1015 Lausanne, Switzerlandb Technology and Society Laboratory/Swiss Federal Laboratories for Materials Testing and Research, CH-8600 Dbendorf, Switzerland

c Center for Risk Science and Communication, Department of Environmental Health Sciences, School of Public Health, University of Michigan, Ann Arbor,

Michigan 48109, USA

Received 28 October 2005; accepted 1 March 2006

Abstract

The environmental performance of presently operated GSM and UMTS networks was analysed concentrating on the environmental effects ofthe End-of-Life (EOL) phase using the Life Cycle Assessment (LCA) method. The study was performed based on comprehensive life cycleinventory and life cycle modelling. The environmental effects were quantified using the IMPACT2002+ method. Based on technological forecasts,the environmental effects of forthcoming mobile telephone networks were approximated.

The results indicate that a parallel operation of GSM and UMTS networks is environmentally detrimental and the transition phase should bekept as short as possible. The use phase (i.e. the operation) of the radio network components account for a large fraction of the total environmentalimpact. In particular, there is a need to lower the energy consumption of those network components. Seen in relation to each other, UMTSnetworks provide an environmentally more efficient mobile communication technology than GSM networks. In assessing the EOL phase,recycling the electronic scrap of mobile phone networks was shown to have clear environmental benefits. Under the present conditions, material

recycling could help lower the environmental impact of the production phase by up to 50%. 2006 Elsevier Ltd. All rights reserved.

Keywords: Mobile phone networks; Mobile telephony; UMTS; GSM; Life cycle assessment; End-of-Life phase; Environmental impact; Data transfer rate

1. Introduction

The presented study aims at providing in-depth knowledgeon the environmental consequences related to the life cycles ofGSM1 and UMTS2 mobile phone networks, concentrating onthe EOL phase.3 Based on representative forecasts, prognosesof the environmental consequences related to forthcoming

mobile phone network infrastructure are made and recommen-dations to the concerned stakeholders are formulated.

Mobile telephony, presently superseding wired telephony,has become one of the most convenient information exchangetools since the implementation of the GSM standard in the early1990ies. The subscriber numbers rising by hundreds per second(GSMworld, 2005) and the new mobile phone networkslaunched weekly (GSMAssociation, 2004) may help toillustrate this trend. Today GSM technology, modified and

improved by high-speed data transmission techniques (GPRS4,EDGE5), has arrived at a nearly fully developed state. In orderto provide real universal mobile phone access and enabling stillfaster data transfer rates, the UMTS standard has been underdevelopment since 1987 (Hillebrand, 2002) and the firststandard package was adopted in 1999 (ETSI, 1999b). In2004, the first commercial UMTS networks were rolled out in

Environment International xx (2006) xxx xxx

+ M

EI-01451; No of Pages 20

www.elsevier.com/locate/envint

Corresponding author. Technology and Society Laboratory/Swiss FederalLaboratories for Materials Testing and Research, CH-8600 Dbendorf,Switzerland. Tel.: +41 44 823 48 30; fax: +41 44 823 40 42.

E-mail address: [email protected] (W. Scharnhorst).1

Global System for Mobile communication. GSM belongs to the so-calledsecond generation (2G) of mobile phone networks.

2 Universal Mobile Telecommunication System. UMTS networks belong tothe so-called third generation (3G) of mobile phone network.3 End-of-Life phase.

4 General Packed Radio Service. GSM-GPRS networks belong to the so-called second and a half generation (2.5G) of mobile phone networks.5 Enhanced Data rates forGlobal Evolution. GSM-EDGE networks belong to

the so-called second and a half generation (2.5G) of mobile phone networks.

0160-4120/$ - see front matter 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.envint.2006.03.001

ARTICLE IN PRESS
mailto:[email protected]://dx.doi.org/10.1016/j.envint.2006.03.001http://dx.doi.org/10.1016/j.envint.2006.03.001mailto:[email protected]


2/20

Western Europe. For the future, consultants expect worldwidesuccess of the UMTS technology similar to the success of theGSM technology (Delpho, 2005; Schullitz, 2001).

Although mobile telephony provides undeniably usefulservices, it can cause relevant environmental impacts, e.g.through the dramatically growing amount of electronic scrap,

inefficient energy management during its operation and servicetimes of, in particular, mobile phones that are still too short. Thechange-over and the associated competition between the GSMand the UMTS technology will further exacerbate these

problems. In order to reduce the environmental impacts ofelectric and electronic equipment and of electronic scrap, theEuropean Union has adopted regulations to ban hazardoussubstances from electronics (CEC, 2003a), and to increase therecycling rate of electronic scrap (CEC, 2003b). Likewisesubstantial efforts have been undertaken by the telecommuni-cation industry. For example, today the subscriber contracts lastlonger, typically about two years. Correspondingly, the service

time of a mobile phone has been expanded to 1.5

2 yearscompared with 0.51 year in 2001 (Swisscom, 2005b).In the context of the qualitative and quantitative analysis of

the environmental consequences of large technical systems, suchas mobile telephone equipment, LCA6 has been recognised as a

powerful tool. This method provides a framework with which tolocalise potentials to improve the environmental performance ofmobile phone networks and components.

Recent LCA studies as well as experimental analyses haveinvestigated the environmental effects related to

electronic elements contained in network components (Uryuet al., 2003),

separate mobile phone network components (Fishbein, 2002;Grunewald and Gustavsson, 1999; RANDA-GROUP, 2000;Scharnhorst et al., 2005b; Tanskanen and Takala, 2001), and

entire mobile phone networks (Faist-Emmenegger et al.,2004; Malmodin et al., 2001; Pehrsson and Hedblom, 2005;Scharnhorst et al., 2005a; Weidman and Lundberg, 2001).

Most of the LCA studies arrive at the conclusion that the usephase dominates the overall environmental impact of thenetworks and/or the components. The other phases mostlyseem to have a minor (production phase) or negligible (EOL

phase) environmental impact. The contributions of the separate

network components to the total network impact are contro-versially debated and study results are differing. Only a veryfew studies have considered the upcoming UMTS standard

based on data that thus far have been deficient.Thus, although, LCA studies in general provide substantial

environmental know-how, for assessment of large technicalsystems, such as mobile phone networks, they are often subjectto weaknesses resulting in possibly biased results:

the analysed systems are highly complex, and the analysed systems are modelled in an oversimplified way,

e.g. the EOL phase is not modelled in a comprehensive way.

Bearing these issues in mind, an LCA study was performedconcentrating on:

i.) the comparison of the environmental performance of aGSM network (corresponding to the ETSI7 standard

package Release 1997 (ETSI, 1996)) and a UMTS

network (corresponding to the 3GPP8 standard package (R'99)9 (ETSI, 2002c)) as presently operated inSwitzerland,

ii.) a realistic analysis of the environmental consequences ofthe EOL phase of both network types,

iii.) a prognosis on the environmental performance of GSMnetworks technically modified for accelerated datatransfer (using GPRS and EDGE) and of UMTS networkslikewise modified (corresponding to the 3GPP standard

packages (R'04)10 (ETSI, 2003c) and (R'06)11 (ETSI,2005), and

iv.) a sensitivity analysis of the key influencing parameters:

number of subscribers and total data download volume.

This paper compiles the results representative for GSM andUMTS (R'99) technology in Switzerland in 2004. It alsodocuments prospective results for UMTS networks complyingwith the upcoming standard packages (R'04) and (R'06). In

particular, it aims to address the following issues:

a) When comparing GSM and UMTS, which type of networkperforms better environmentally?

b) When the EOL phase of networks is properly modelled,which life cycle phase dominates the total environmentalimpact of the networks?

c) When the networks are modelled according to the relevantstandards and including all of the major network compo-nents, which component of the networks dominates the totalenvironmental impact?

d) Is processing of electronic scrap and the production ofsecondary raw materials in the EOL phase more environ-mentally relevant than the production of primary rawmaterials in the production phase?

e) When high-speed data transfer techniques for GSM (GRPSand EDGE) are included, do these techniques help to lowerthe environmental impact of the GSM network?

f) When UMTS network alterations complying with the future

standard packages (R'04 and R'06) are considered, what willbe the environmental impact of such UMTS networks?

The presented LCA study was performed and the paper isstructured in compliance with the ISO 14040 series (ISO 1998a,

b) into the following sections: goal and scope definition, lifecycle inventory, life cycle impact assessment and resultsinterpretation. The paper is complemented by a sensitivity

6 Life Cycle Assessment.

7 European Telecommunications Standards Institute.8 3rd Generation Partnership Project.9 Release 1999.

10

Release 2004.11 Release 2006.

2 W. Scharnhorst et al. / Environment International xx (2006) xxxxxx

ARTICLE IN PRESS


3/20

analysis and is completed by a discussion and recommendationsto the stakeholders concerned.

2. Method, goal and scope

2.1. Methodological background

The continuously increasing exploitation of naturalresources and the growing amount of emissions associatedwith, in particular, industrial activities necessitated the imple-mentation of measures to monitor and manage the effects on theenvironment. The Life Cycle Assessment (LCA) methodology

provides a consistent framework aiming at the assessment ofenvironmental aspects and potential impacts associated with a

product/service12 (ISO, 1997). In practice, LCA can assist toidentify environmental aspects of products at different life cyclestages and thus can be used for decision-making in industry orgovernment (ISO, 1997). An LCA typically consists of four

phases: goal and scope definition, life cycle inventory (LCI13

),life cycle impact assessment (LCIA14) and result interpretation(ISO, 1997; Rebitzer et al., 2004). Following the standardsdefined for LCA, in the goal and scope section of the presentedstudy the studied system was defined, the data representativitywas specified, and the impact assessment method wasdetermined (ISO, 1997). Subsequently the life cycle inventory

part was performed for the various processes of the production,use and EOL phases. The environmental data of the separatenetwork components (i.e. the resource consumptions and theemission releases) were inventoried in compliance with theearlier defined scope of the study. Processes that did not fit intothe previously defined study scope were not inventoried. The

inventory step also included the assembly of the network model(life cycle modelling) and the allocation of material and energyflows (ISO, 1997, 1998a). In compliance with the ISO standardsfor LCA, which require a broad coverage of impact categories(Pennington et al., 2004), in the impact assessment section ofthe presented study the environmental impacts of the networkand/or its components were calculated by assigning impactscores to the various resource consumptions and emissionswhich had been compiled in the life cycle inventory part (ISO,1997, 1998b). In the final section the results of the impactassessment were interpreted, a sensitivity analysis was

performed for the obtained results, conclusions were drawn

and recommendations were formulated for the attention of theconcerned stakeholders.

2.2. System description

2.2.1. Functional unit and reference flow

The environmental impact of a product or service is relatedto the functionality it provides. In order to cover the key

functionalities of GSM and UMTS mobile phone networks(voice and data transmission) the transmission of data (i.e.speech and non-voice applications) in bit from a mobile phonevia the mobile phone network was selected as functional unitand 1 bit of traffic data transmitted was selected as the referenceflow (ISO, 1998a). The selected functional unit enables a

common analysis of environmental effects associated with i.)voice transmission (typically measured in tax minutes) and ii.)non-voice, i.e. data transmission (typically measured in bitstransmitted). This basic functional unit has the advantage thatother more application-oriented functional units can bederived from it. For example, one tax minute of phoneconversation requires the transmission of about 576 kbit inthe case of GSM and 732 kbit in the case of UMTS.

2.2.2. Data requirements

The following requirements were set and the data should berepresentative for:

Western Europe with respect to services, frequencies, datatransfer rates, etc., and for Switzerland with respect tonetwork load,

the year 2005 with respect to GSM (including GPRS andEDGE) and UMTS (R'99) networks and for 2006/07 withrespect to UMTS (R'04, R'06) networks, and

Western Europe and 2005 with respect to EOL treatment.

2.2.3. System boundaries

The system under study encompasses all life cycle phases ofa representative GSM, GSM-GPRS and -EDGE and UMTS(R'99) networks as well as of UMTS networks complying with

the forthcoming standard packages (R'04 and R'06).The production phase of any network or its components

begins with the extraction of ores and energy carriers, itincludes the fabrication of base materials such as metalalloys and plastics, and ends as the network componentassembly is finalised, i.e. as the electronic components andthe supporting structural elements (e.g. PWBA15, frames,casings, etc.) are assembled (Scharnhorst et al., 2005a). Thefinal assembly of network components was not included inthe studied system as other studies have proven that thisstage is of minor environmental relevance (Faist-Emmeneg-ger et al., 2003; RANDA-GROUP, 2000; Weidman and

Lundberg, 2001). No environmental data were inventoriedfor these processes.

The use phase follows the production phase. In the case ofmobile phone networks, it includes the network installation andcontinues with the operation of the network components. In

principle, this phase also includes maintenance and repairservices as well as periodical software updates (Scharnhorst etal., 2005a). However, the latter two stages (i.e. maintenance andrepair services) were not included in the presented study due totheir proven low environmental relevance (Faist-Emmeneggeret al., 2003).

12 According to ISO-terminology the term product does also services ISO:ISO 14040: kobilanz Prinzipien und allgemeine Anforderungen,International Organisation for Standardisation Brussels. 1997.13

Life Cycle Inventory.14 Life Cycle Impact Assessment. 15 Printed Wiring Board Assembly.

3W. Scharnhorst et al. / Environment International xx (2006) xxxxxx

ARTICLE IN PRESS


4/20

The life cycle of the mobile phone networks ends with theEOL phase. This phase begins with the dismounting of thedevice to be replaced. Thereafter a more or less sophisticated

pre-processing of the electronic scrap follows. Subsequently,thermal EOL treatment is applied in order to recover energy and

precious metals. The recovered materials are recycled and

energy is reused. For the studied system, it was assumed that allsecondary materials and all energy are reused for manufacturingof new network components (closed-loop). It was not assumedthat any secondary raw materials or energy is used in other lifecycle systems. The residuals of the EOL phase are finallystabilised and landfilled. In the case of the presented study, theenvironmentally most favourable EOL scenario, determinedearlier (Scharnhorst et al., 2005a), was selected to representstate-of-the-art processing of electronic scrap in the EOL phase(Fig. 1).

2.2.4. Allocation

The foreground system was modelled based on physicalproperty allocation when there was a physical causality (Ekvalland Finnveden, 2001). For example, the usage of a shredder isallocated to the mass of the treated scrap.

For the background system, the allocation rules as applied inthe ecoinvent-database were adopted (Frischknecht et al.,2004). According to these guidelines, multi-output unit

processes are inventoried in the database prior to allocation.Thereafter the mass and energy flows are allocated to each co-

product generated by a multi-output process by application ofallocation factors (Frischknecht et al., 2004). The unit processdata sets obtained by this procedure were then used to model the

background system.

With respect to precious metals, such as gold, silver,tungsten, etc., no appropriate production data sets wereavailable. These processes were approximated as single-output

processes, i.e. no co-products were taken into account, based on

publicly available information (ISO, 2005). The approximationas single-output processes makes an allocation superfluous.

The EOL phase was modelled including system expansion(Scharnhorst et al., 2005b). Credits, i.e. numerically negativeenvironmental impacts, were considered in the EOL phase withrespect to the production of secondary raw materials and the

energy recovery, substituting primary materials and energy fromprimary sources respectively (BDSV, 2001; HELCOM, 2002;Lehner, 2001; VSSV, 2005).

2.2.5. Impact assessment

The IMPACT2002+ method (Jolliet et al., 2003) was used inorder to determine the environmental impacts related toresources consumed and emissions released during the lifecycle of the mobile phone networks. This method, based on theIMPACT2002 model (Pennington et al., 2005), comprises 14midpoint categories and four damage categories: human health,ecosystem quality, climate change and resource consumption.

The method links the input and output flows inventoried for acertain object to derive effect scores of specific referencesubstances. The scores are finally linked to the damagecategories to yield a measure of the environmental impact ofa product or service. The CML2001 method (Guine et al.,2001) was used in order to evaluate the results obtained with theIMAPCT2002+ method.

2.2.6. Interpretation and weighting

The results of the impact assessment were interpreted foreach damage category. No additional normalisation andweighting and thus no additional aggregation were applied.

2.2.7. Review process

The entire study in all its subsequent sections was subject tointernal review (performed by members of the participatinginstitutions).

Dismounting

Collection/ Take back

Pre-separation

Dismantling

Shredding

Fractionation

Material/ Energyrecovery

Incineration

Secondaryraw

materials

Landfilling

EOL Phase

Yes No

MobileTelephony

USE Phase

NetworkNetwork

Powersupply

Primaryraw

materials

Assemblyof printed

wiringboards

Powersupply

PROD Phase

No

No

NoYes

Yes

EnergyNon-voiceData

Voice data(telephony)

Fig. 1. System boundaries of the life cycle phase of the investigated mobile phone networks investigated.


ARTICLE IN PRESS


5/20

2.3. Description of mobile telephone technologies and

techniques considered

2.3.1. GSM

The GSM network analysed reflects GSM technology ascurrently operated in Switzerland. It complies with the GSM

standard packages defined and adopted in 1996 for the basicnetwork architecture (ETSI, 1996), in 1999 for the interface

principles (ETSI, 1999a), and in 2000 for the radio transmissionprinciples (ETSI, 2000a,c).16 In GSM a combination of theFDMA17/TDMA18 techniques is applied to bridge the airinterface between the MS19 and the BTS20. Independent of thefrequency band deployed (900 or 1800 MHz), 124 trafficchannels are provided for the uplink and downlink direction(Duque-Antn, 2002). Each of these channels can be sharedsimultaneously by eight subscribers. Typically, voice transmis-sion is performed at 9.6 kbit/s (details are compiled in AppendixB). Voice transmission is performed in the circuit switched

mode, i.e. during a phone call a subscriber is permanentlyphysically connected to the mobile phone network. Datatransmission is performed at 9.6 kbit/s in the uplink and thedownlink direction. The same channels are used as for voicetransmission. As in the case with voice, data are processed incircuit switched mode. An analysis of the principle architectureof the GSM network and its modifications (GPRS and EDGE)can be found in Appendix B. A description of the separatenetwork components is available in Scharnhorst et al. (2005a)and in Scharnhorst (2005).

2.3.2. GPRS

This technique was introduced in 1995 (Sanders et al., 2003)

in order to boost the remote data transfer rates of GSMnetworks. Using this service, data (e.g. e-mail, web-documents,etc.) are transmitted at a maximum data rate of 31.2 kbit/s(uplink) and 62.4 kbit/s (downlink) (ETSI, 2001a). Theoreti-cally 171.2 kbit/s are possible but not practiced (Rudolf, 2003).The data are processed in packet switched mode, which allowsfor the faster transfer rates. Information on the modification ofthe physically underlying GSM network (e.g. additionalnetwork components) are compiled in Appendix B and in therelevant literature (ETSI, 2000b, 2003b; Halonen et al., 2003;Sanders et al., 2003).

2.3.3. EDGEThis data transmission technique represents the final step in

the evolution of mobile telephony from GSM towards UMTS,i.e. from 2G to 3G (Halonen et al., 2003). The application of adifferent signal modulation scheme allows for another increasein data transfer rates per traffic channel (ETSI, 2000d, 2003a).Maximal available data rates up to 192 kbit/s (uplink) and

384 kbit/s (downlink) are theoretically possible (Schnabel,2004). Again, this service applies to data transmission and dataare processed in the packet switched mode (details are compiledin Appendix B).

2.3.4. UMTS (R'99)

The UMTS network investigated complies with the UMTStechnology currently practiced in Switzerland. It meets theUMTS standard packages as defined and adopted in 2002 forthe basic network architecture (ETSI, 2002c,e), in 2000 for datatransfer service (ETSI, 2000b), and in 2002 for the interface

principles (ETSI, 2002a,b).16 Radio transmission in UMTS, i.e.the signal exchange between Ue21 and NodeB, is performedusing CDMA22 (in Europe W-CDMA23). The CDMA tech-nique does not foresee any discrete channel separation. Instead,all subscribers use the same frequency spectrum (FDD24:5 MHz in the 1900/2100 MHz band; TDD25: 5 MHz in the1900/2000 MHz band). For the transmission of voice, the so-

called voice service (Banet, 2005b; Banet et al., 2004) at12.2 kbit/s is deployed (details are given in Appendix B).According to the 3GPP standard package (R'99), voicetransmission is performed in the circuit switched mode (Banetet al., 2004). In contrast to the GSM standard, data transmissionin UMTS is packet switched (Banet et al., 2004), i.e. asubscriber is physically connected with the network only aslong as data is transferred. The 3GPP standard package (R'99)foresees the operation of UMTS networks in the FDD and in theTDD mode (ETSI, 2002d). Presently however, most UMTSnetworks are operated in the FDD mode (Grtner, 2005). Thismode limits the theoretical maximum data transfer rates to64 kbit/s (uplink) and 384 kbit/s (downlink) (Banet et al., 2004;

Schnabel, 2003, 2004). Further details on data processing arepresented in the cited literature compiled in Appendix B andoutlined in Scharnhorst (2005). The composition of the UMTSnetwork analysed and its modifications (R'04 and R'06) can befound in Appendix B. A description of the separate componentsis available in Scharnhorst (2005).

2.3.5. UMTS (R'04/R'05)

This first substantial modification of the original UMTSstandard represents an evolution towards: i.) packet switchedvoice transmission and ii.) the implementation of the TDDmode for high-speed data transfer (Banet, 2005a; ETSI,

2003c; Holma and Toskala, 2004). Using TDD, maximaldownload data transfer rates of 1920 kbit/s and upload ratesof 960 kbit/s are possible (Banet et al., 2004). Although

packet switched voice transmission is optionally available, itis assumed that networks complying with these standard

packages still perform voice transmission in the circuitswitched mode (Banet, 2005a).

16 Mobile phone standards are subject to permanent updates. To take thepresent conditions into account, the latest standard versions were adopted.17 Frequency Division Multiple Access.18 Time Division Multiple Access.19 Mobile Station (GSM mobile phone). For convenience throughout the

paper, except in the diagrams, MS will be termed mobile phone.20 Base Transceiver Station.

21 User equipment (UMTS mobile phone). For convenience throughout thepaper, except in the diagrams, Ue will be termed mobile phone.22 Code Division Multiple Access.23 Wideband-CDMA.24

Frequency Division Duplex.25 Time Division Duplex.


ARTICLE IN PRESS


6/20

2.3.6. UMTS (R'06)

This most recently adopted standardisation package (ETSI,2005) in particular aims at a further increase in data transferrates using HSDPA26 and HSUPA27 (ETSI, 2003d, 2004).Uplink data transfer rates up to 5800 kbit/s (Ihlenfeld, 2005) anddownlink rates up to 14400 kbit/s are envisaged (Anonymous,

2005a; Holma and Toskala, 2004; ETSI, 2001b). This standardpackage concentrates on the simultaneous packed switchedprocessing of voice and data (All-IP) (Holma and Toskala,2004).

3. Life cycle inventory system modelling

3.1. Network modelling

A basic GSM network was modelled in compliance withthe above addressed standards and for Swiss conditions basedon the most recent statistics (BAKOM, 2004a; Orange, 2004;

Sunrise, 2005; Swisscom, 2005a). Likewise, the UMTS(R'99) network was modelled based on current statistics(Scholz, 2005). Future UMTS networks complying with theforthcoming standards (R'04 and R'06) were modelled basedon recent forecasts (PhoneContent.com, 2004; Schullitz,2001). For each of the networks studied, the technicallyfeasible maximum data transfer rates were adopted. In orderto provide higher data transfer rates, GPRS and EDGErequire the bundling of traffic channels. It was assumed thatfour traffic channels are combined to get one GPRS datachannel, and eight traffic channels to get one EDGE datachannel. Also, it was assumed that a BTS covers threesectors. In UMTS different modulation and access schemes

provide higher data transfer rates. In order to alter thecapacity of a NodeB, the number of covered sectors and cellsand the amount of power amplifiers installed at the NodeBsites can be varied. In the presented study, it was assumedthat a NodeB of a UMTS (R'99) network covered threesectors and three cells. In the case of UMTS (R'04) and(R'06) networks three sectors and six cells (two cells persector) were assumed to be covered. Further details arecompiled in Appendix B.

The technical specifications of the separate networkcomponents were compiled from original manufacturer datasheets (sources are given in Appendix C). Information on the

basic network component materials were adopted fromGoosey and Kellner (2002), Ludwig et al. (2003), Motorola(2005), Scharnhorst et al. (2005a) and from the networkcomponent manufacturers listed in Appendix C. Supplemen-tary information on the network architecture was compiledfrom relevant books (Banet et al., 2004; Bekkers and Smits,1997; Benkner and Stepping, 2002; Duque-Antn, 2002;Eberspcher et al., 2001; Halonen et al., 2003; Rudolf, 2003;Sanders et al., 2003; Schnabel, 2003; Steele et al., 2001).Information on the operation modes of the networks wasobtained from component manufacturers (Grtner, 2005).

The data used to model the basic and the modified GSMnetworks (GPRS and EDGE) as well as the basic UMTS(R'99) network refer to Western European conditions in 2005.The data used to model the evolved UMTS networks (R'04)and (R'06) refer to tentative forecasts of European networkoperators and are valid for Western European conditions

(ERICSSON, 2005; Janssen, 2005; Nokia, 2003). All networkcomponent data comply with the respective standardsmentioned above.

In order to simulate the different operating conditions, thefollowing parameters were introduced:

Seasonal parameter(SP1)

Varies the seasonal conditions; the energyconsumption of the network components is alteredaccordingly.

Traffic parameter(TP)

Adjusts the energy consumption of the radio andcore network components according to thesubscriber load.

Data transfer parameter

(DTP)

Adjusts the data transfer rates according to the

technically feasible maximum data rates. Subscriber parameter

(SP2)Modifies the number of subscribers according tosubscriber forecasts.

Download volumeparameter (DVP)

Varies the total data download volume persubscriber according to hypothetical estimates.

Table 1 documents selected present average conditionsadopted for the mobile phone networks and the parameters thatwere used. The detailed operating conditions for the networksare compiled in Appendix B.

3.2. Life cycle modelling

The life cycles of the mobile phone networks were dividedinto the three phases: production, use and End of Life treatment.

The production phase was modelled as defined in the systemboundaries paragraph (see Section 2.1) and information on theenergy consumption in the PWBA manufacturing were adopted

26

High Speed Downlink PacketAccess.27 High Speed Uplink Packet Access.

Table 1Selected average operating network conditions representing present conditionsand parameters

Technical network parameters Networks G SM UMTS(R'99)

Model

parametersData rate [kbit/s] DTP 9600 384,000MS (Mobile station)/Ue

(User equipment) []SP2 6,188,793 70,000

Phone call [s/MSyear] 87 600 87,600Total data download volume

[Mbit/MSyear]DVP 0.7 564.0

BTS (Base TransceiverStation)/NodeB []

SP1, TP,DTP, DVP

6800 3465

BSC (Base Station Controller)/RNC(Radio Network Controller) []

" 50 23

MSC (Mobile Switching Centre) [] " 34 15SGSN (Serving GPRS

Support Node) []" - 23

GGSN (Gateway GPRS Support Node) [] " - 23

BBC (Back Bone Cable network) [km] " 95,000 66,000


ARTICLE IN PRESS


7/20

from Kincaid and Geibig (1998). The production phase wasmodelled adopting process data compiled in the ecoinvent-database (ecoinventCentre, 2003). Transport process data wereadopted from the GaBi4-software (IKP and PE, 2003). Littleinformation was available on the production of metals, in

particular precious metals, thus necessitating an explicit

approximation of the production processes for: antimony,arsenic, beryllium, bismuth, brome, cadmium, europium,gallium, germanium, gold, indium, indium-tin-oxide, lithium,ruthenium, silver, thallium and tungsten.

In the use phase, only the energy consumed to operate and toaerate the network components was included. Seasonal weatherconditions that influence the energy consumption of thenetwork components were included by varying the energyconsumption for the HVAC28 of the network components. Peakenergy consumptions of the network components wereaddressed by using a traffic parameter (TP). The parameterwas derived from the total annual mobile phone traffic

(BAKOM, 2004a). Energy consumption data of the networkcomponents considered were adopted from the manufacturerslisted in Appendix C, from service personnel (Hausammann,2005) and from one network operator (Swisscom, 2004).Mobile phone power consumption data were estimated fromreports (AFU, 2004; Stromtip.de, 2000) and the manufacturerslisted in Appendix C. The energy supply processes weremodelled adopting data sets compiled in the ecoinvent-database(ecoinventCentre, 2003).

The EOL phase was modelled in detail according to ascenario identified earlier representing the environmentally

preferable EOL option (Scharnhorst et al., 2005a). This EOLalternative includes the dismounting and collection of the

network components as well as a rough dismantling. Thesesteps are followed by state-of-the-art thermal processing inorder to recover metals and energy. The EOL treatment iscompleted by the incineration (stabilisation) of the residuals and

by dumping the stabilised residuals in landfill sites. For allmobile phones it was assumed that 20% of the discarded phonesare not processed in recycling facilities, but are directly

processed in MSWI29 plants. Net recycling rates of 75%(aluminium and steel) and of 95% (precious metals such asgold, palladium, silver) were assumed for metal recycling(BDSV, 2001; HELCOM, 2002; Lehner, 2001; VSSV, 2005).30

For the MSWI plant, it was assumed that the efficiency of the

filter units located downstream from the incineration facilitieswas 90% (Farrell, 2000). For the EOL treatment, transfercoefficients31 and fractions for the incineration stage weredirectly estimated based on experimental measurements(Scharnhorst et al., 2005c). Coefficients and fractions for the

landfilling stage were estimated based on physio-chemicalproperties (Scharnhorst et al., 2005b). Information on themechanical and thermal EOL treatment was partly obtainedfrom recyclers (BOLIDEN, 2002; Stengele, 2004) and partlyfrom literature (Ludwig et al., 2003; Scharnhorst et al., 2005b).Technical specifications were adopted from relevant data sheets

(Berzelius, 1993; BckmannGmbH, 2001a,b,c; Weyhe, 2004).All foreground EOL processes, e.g. the infrastructure, werespecifically modelled. All background EOL processes (e.g.fabrication of most of the base materials of which, for instance,the EOL infrastructure consists) were modelled adopting datasets from the ecoinvent-database (ecoinventCentre, 2003).Transport process data were adopted from the GaBi4-software(IKP and PE, 2003). Administration processes are notconsidered in any of the life cycle phases.

All data adopted are applicable under the data requirementsset out above.

4. Life cycle impact assessment results (IMPACT2002+)

In this section, the environmental performance is presentedof the GSM and UMTS networks investigated with respect tothe functionality of the networks (i.e. related to the datatransmission). All results presented represent relative environ-mental impacts per functional unit. The absolute environmentalimpacts are addressed in the discussion section.

4.1. Resource depletion

This damage category is dominated by the total environ-mental impact score of the UMTS (R'99) network (71%32).

Comparing the resource depletion effects of the two networkstandards reveals that GSM networks (basic configuration33)under present conditions perform better than UMTS (R'99)networks (factor 8). Upgrading from UMTS (R'99) to (R'04and R'06)34 will lead to lower environmental profiles per bitcompared with UMTS (R'99) (factors 835 and 2135). Theupgrade of UMTS network technology results in environmental

profiles close to that of GSM networks equipped with theEDGE-technique (factor 0.335 and 0.835 in the case of UMTS(R'04) and (R'06) respectively). However, the installation ofadditional network infrastructure (NodeB, RNC, etc.) as well asthe increased energy consumption in the use phase moderately

increases the absolute total annual consumption of the network(see Section 5) and limits the environmental benefit perfunctional unit.

The total environmental impacts of the basic standardnetworks (i.e. GMS and UMTS (R'99)) are dominated by theuse phase (80%36 and 56%36) (Fig. 2). Key processes are the

28 Heating, Ventilation, Air Conditioning.29 Municipal Solid Waste Incineration plant.30 These net recycling rates are representative for the pure recycling process

and take no losses during scrap sorting and collection into account. In case oflarge network components this loss was assumed to be not relevant. In case ofsmall network components, such as mobile phones, this loss was taken intoaccount by assuming that 20% of the mobile phones are discarded without any

recycling.31 These coefficients are used to quantify the volatilisation of e.g. metals.

32 Total impact of all networks.33 i.e. no GPRS and no EDGE.34 Forecasted subscriber numbers and increased total data download volumes

per subscriber are taken into account.35 Additional environmental impacts due to the expansion of network

infrastructure and due to the increased power consumption of the network

components are taken into account.36 Share of total impact of the respective network.


ARTICLE IN PRESS


8/20

energy consumption by the NodeB (89%37) and the BTS(79%37). The uranium depletion related to the Swiss electricitygeneration contributes to the impact score of the use phase

(64%36 in the case of the GSM network and 72%36 in the caseof the UMTS (R'99) network). The production phasecontributes only to a limited extent to the total environmentalimpact of a UMTS (R'99) network (20%36), which is mainlydue to the generation of electricity for the production of primaryaluminium (54%37), and contributes dominantly to the impactscore. The production phase of the GSM network accounts for44%36 of the total impact score and is dominated by the energyintensive manufacturing of PWBA for mobile phones and BTSracks (71%37). The environmental benefit related to therecycling and manufacturing of secondary raw materials inthe EOL-phase is restricted in terms of resources both for the

GSM and the UMTS networks (

5%36

and

6.0%36

respectively).

4.2. Climate change

The total environmental impact of the UMTS (R'99)network is the highest for this damage category and resultsare very similar to the resource depletion category (70.5%38)(Fig. 3). Comparing the climate effects of the two networkstandards shows that GSM networks (basic configuration39)

under the given initial conditions perform better than UMTS(R'99) networks (factor 8). Upgrading from UMTS (R'99) to(R'04 and R'06)40 will lead to lower environmental profiles

compared with UMTS (R'99) (factor 841 and 2241). The UMTSnetwork upgrade again results in environmental profiles close toGSM networks deploying the advanced EDGE-technique(factor 0.341 and 0.941 in the case of UMTS (R'04) and(R'06) respectively).

As before, the use phase dominates the total climate changescore in each of the UMTS networks (68%42 for a UMTS(R'99) network, 72%42 and 69%42 in the case of a UMTS(R'04) and (R'06) network respectively). The impact scores ineach network configuration are attributable to the CO2emissions associated with the generation of electrical energysupplied to operate the NodeB (88%43 in the case of a UMTS

(R'99) network). CO2 emissions related to the energy intensiveproduction of primary aluminium for NodeB racks (27%43) andthe manufacturing of the PWBA used in NodeB racks andmobile phones (32%43) account for the impact score in the

production phase. The recycling of electronic scrap in the EOLphase can account for a reduction of the total environmental

37 Share of total phase impact of the respective network.38

Total impact of all networks.39 i.e. no GPRS and no EDGE.

Fig. 2. Resource depletion damage category: Life cycle phase specific and total impact scores of the GSM network (basic, GPRS and EDGE-technique) the UMTSnetworks (R'99, R'04, R'06). The abbreviations of the network components displayed in this diagram and all following diagrams are explained as follows. MS LIIO:mobile phones with rechargeable lithium ion batteries (GSM). MS NIMH: mobile phones with rechargeable nickel-metal-hydride batteries (GSM). BTS: BaseTransceiver Station (GSM). BSC: Base Station Controller (GSM). MSC: Mobile Switching Centre (GSM/UMTS). SGSN: Serving GPRS Support Node (GSM/UMTS). GGSN: Gateway GPRSSupport Node(GSM/UMTS). BBC: BackBone Cable network of the mobile phone network (GSM/UMTS). Ue LIIO: mobile phoneswith rechargeable lithium ion batteries (UMTS). RNC: Radio Network Controller (UMTS).

40 Forecasted subscriber numbers and increased total data download volumesper subscriber are taken into account.41 Additional environmental impacts due to the expansion of network

infrastructure and due to the increased power consumption of the networkcomponents are taken into account.42

Share of total impact of the respective network.43 Share of total phase impact of the respective network.


ARTICLE IN PRESS


9/20

impact score of 16%44 and it can halve the environmentalimpact score of the production phase. In particular energysavings due to the recovery of aluminium (61%45) and silver(18%45) contribute to this reduction.

In contrast to the UMTS networks, under the givenconditions, the total environmental impact score of the GSM

networks is dominated by the production phase (52%44). Theimpact score of this phase is dominated by CO2 emissionsassociated with the energy intensive PWBA manufacturing ofmobile phones and BTS racks (65%45). The use phase isdominated by CO2 emissions associated with the energysupplied to operate the BTS (78%45). The EOL phase canreduce the climate change score by 11.6%44 a quarter ofthe impact score of the production phase. Again, the energysavings due to recycling of silver (54%45) and aluminium(36%45>) account for this reduction.

4.3. Human health

Under the defined conditions, the UMTS (R'99) networkdominates this damage category and accounts for 70%46 of thetotal environmental impact of all networks (Fig. 4). Networkscomplying with the GSM standard perform better comparedwith the basic UMTS (R'99) network (factor 747 in the case of a

basic GSM network, factor 1347 in the case of a GSM-GPRSnetwork, and factor 2547 in the case of a GSM-EDGE network).

Future UMTS networks (R'04 and R'06) will have anenvironmental performance close to that of GSM networksdeploying the most recent EDGE-technique (factor 0.347 in thecase of UMTS (R'04), and factor 0.947 in the case of UMTS(R'06)).

Under the given initial conditions, the use phase of a UMTS

(R'99) network accounts for a majority of the network's totalenvironmental impact score (52%44), in particular due to theenergy supplied to operate the NodeB. Primary and secondary

particles are the main impacting sources, due to SO2, NOx andparticle emissions to air (25%45, 19%45, 29%45). The impactscore of the production phase is dominated by the production of

primary aluminium for the NodeB racks (22%45) and theproduction of primary lead for the NodeB back-up batteries(20%45). The energy intensive PWBA manufacturing (12%45),and the primary steel production for the NodeB racks and the

primary palladium production for the PWBA (each 7%45) arealso significant. Again, primary and secondary particles are the

main sources for the impact score of this phase. The EOL phasecan account for a reduction of the total environmental impact by29%48 and it can lower the environmental impact of the

production phase by a factor of 1.6. In particular the recycling ofaluminium (35%49), steel (18%49) and palladium (11%49) of the

NodeB racks can account for the reduction of environmentalimpact score.

In contrast to the UMTS networks, the production phasedominates the human health impact score of GSM networks(63%48). This is due to primary and secondary particlegeneration linked with the energy intensive manufacturing of

Fig. 3. Climate change damage category: Life cycle phase specific and total impact scores of the GSM network (basic, GPRS and EDGE-technique) the UMTSnetworks (R'99, R'04, R'06).

44 Share of total impact of the respective network.45 Share of total phase impact of the respective network.46 Total impact of all networks.47 Additional environmental impacts due to the expansion of network

infrastructure and due to the increased power consumption of the networkcomponents are taken into account.

48

Share of total impact of the respective network.49 Share of total phase impact of the respective network.


ARTICLE IN PRESS


10/20

PWBA for mobile phones and BTS racks, as well as theproduction of primary palladium, silver50 and aluminium for theBTS racks (26%49,15%49,10%49 and 7%49). Additional effectsare attributable to the manufacturing of rechargeable nickel-metal-hydride batteries (15.9%49). SO2, NOx and particleemissions to air contribute to the impact of this phase

(48.2%49, 16.6%49, 24.0%49). The use phase of a basic GSMnetwork accounts for 37%48 of the human health impact scoremainly due to the operation of the BTS (79%48) and the mobile

phones (17%48). The recycling of network scrap in the EOLphase can account for a reduction of the network's totalenvironmental impact (23%48) and it can nearly halve theenvironmental impact score of the production phase due torecovery of secondary palladium (39%49), aluminium (19%49),and steel (13%49).

4.4. Ecosystem quality

Under the given conditions, the UMTS (R'99) networkshows the highest total environmental impact score onecosystem of all networks analysed (57.9%51) (Fig. 5), althoughassessment uncertainty is higher for this impact category.Again, GSM networks of any configuration perform better thana UMTS (R'99) network (factor 452 in the case of a basic GSMnetwork, factor 752 in the case GPRS is deployed and factor1352 if EDGE is used). Under future conditions (increased totaldata download volumes per subscriber and increased subscriber

numbers) UMTS networks will have an environmentalperformance close to or better than that of GSM networksusing the EDGE-technique (factor 0.552 and 1.252 in the case of(R'04) and (R'06) respectively).

The total impact on ecosystems of a UMTS (R'99) networkoperated under the defined initial conditions is dominated by the

use phase (70%48), mainly due to the energy consumed by theNodeB. The main emissions are, in decreasing order, copper tosoil, zinc to water and copper to air (generated in the electricitygeneration by means of hard and brown coal (Frischknecht andFaist-Emmenegger, 2003; Rder et al., 2004)). The impact ofthe production phase is dominated by the fabrication of primaryaluminium for the NodeB racks (46%53). Particularly, emissionsof aluminium to water (released from landfilled redmud of thealuminium oxide production (Althaus et al., 2004)), zinc to air(released as zinc dust in the production of primary lead for the

back-up batteries (Althaus et al., 2004)) and zinc to soil(released from landfilled by-products of the lead production

(Althaus et al., 2004)) contribute to the impact score of thisphase.

Under the given conditions, the environmental profiles of theGSM networks are completely different from those of theUMTS networks and from previous impact categories.Dominating life cycle phase is the EOL phase (68%54) withhigh impact scores attributable to long-term emission of copperand nickel to soil from dumped incineration ashes of mobile

phones.55 Environmental benefits account for less than 8%54

Fig. 4. Humanhealth damage category: Life cycle phase specific and total impact scores of the GSM network (basic, GPRS and EDGE-technique) the UMTS networks(R'99, R'04, R'06).

53 Share of total phase impact of the respective network.

50 Both metals are important materials in PWB.51 Total impact of all networks.52 Additional environmental impacts due to the expansion of network

infrastructure and due to the increased power consumption of the networkcomponents are taken into account.

54 Share of total impact of the respective network.55

20% of the mobile phones were assumed to be directly incinerated in aMSWI and then referred to a landfill.


ARTICLE IN PRESS


11/20

and are attributable to the recycling of aluminium from BTSracks and of lead from BTS back-up batteries. The use phaseaccounts for only 22%54, mainly linked to the operation of theBTS. The impact of the production phase on ecosystems is lowcompared with the other phases (


12/20

complementing the above LCA analysis, was performed for aGSM-EDGE network and a UMTS (R'04) network57 in orderto demonstrate the improvement capabilities of the environ-mental performance of UMTS networks and the following

parameters were varied (Table 2).Both the absolute overall yearly network performances and

the performances per bit transferred are analysed. Theenvironmental performance of a GSM-EDGE network under

the conditions as adopted for the LCA study (scenario 0) wasselected as the reference and set to 100%.For both networks, an increase in subscribers leads on the

one hand to a moderate increase in the absolute overall yearlyimpact of the network of about 10% to 20% when doublingthe number of subscribers (R04_0 to R04_1 in Fig. 6). This isdue to the additional number of mobile phone produced andthe increased energy consumption of phones and networkinfrastructure during use phase. On the other hand, the impact

per bit strongly decreases when subscribers increase (R04_0to R04_1 in Fig. 7: about 40% when the number ofsubscriber doubles). Similarly, an increasing downloadvolume per subscriber leads to a slight increase of 5% to

10% in the overall yearly network consumption whendoubling the transferred volume ((R04_1 to R04_1a) in Fig.6) and to a further strong reduction in consumption per bit(Fig. 7).

The results show that under the initial conditions(scenario 0) a UMTS (R'04) network has a slightly loweroverall environmental impact per year than a GSM-EDGEnetwork (Fig. 6), due to the low number of subscribers.This inefficient network load of UMTS (R'04) networksresults in a higher environmental impact per bit (Fig. 7). Asdiscussed above the environmental impact per bit stronglydecreases with the increase in subscribers of UMTS (R'04)

that also causes a decrease in GSM-EDGE subscribers(scenario 1). Therefore, the relative environmental perfor-mance of the GSM-EDGE network per bit worsenscompared to the reference situation (Fig. 7). The annualenvironmental impact however, is reduced (Fig. 6). Whenincreasing the total download volume per subscriberadditionally to the increase in subscribers (scenario 1(a)),

or when doubling once more the number of subscribers forUMTS (R'04) (scenario 2), then the relative environmentalimpact of UMTS (R'04) networks is reduced to that of thereference GSM-EDGE network (Fig. 7). A further increasein the total download volume per subscriber (scenario 2(a)),leads to another dramatic reduction in the relativeenvironmental impact at about half of the reference GSM-EDGE network (Fig. 7). It leads again to a moderate

increase in the total annual impact score of the UMTSnetwork (Fig. 6). This increase is partly compensated by thefurther decrease in the GSM-EDGE yearly networkconsumption.

As obvious from the above results from an absolutepoint of view, UMTS networks per se perform environ-mentally worse than GSM networks when it comes toGSM-like full geographic coverage (Fig. 6). This is in

particular attributable to the significantly higher energydemand of the NodeB racks (when operated under fullload a NodeB rack can consume up to 6 kWh(Hausammann, 2005) compared with up to 1.3 kWh ofa BTS rack (SiemensAG, 2000)). Secondly, in order to

cover a geographical area equal to that which a GSMnetwork can cover, up to 30% more NodeB will berequired58 (Hugentobler, 2000). It is obvious that, whencomparing GSM and UMTS simply based on mobiletelephony (i.e. the transmission of speech), UMTS againwill perform environmentally worse due to the abovementioned technological aspects and due to the fact thatthis additional standard is not required, as the service ofmobile telephony is already sufficiently covered by GSM.However, an absolute environmental consideration of two

products, such as GSM and UMTS networks, providingsimilar functions but using different techniques, does not

take into account the added functional value UMTSnetworks can provide. Looking for example at the datatransfer (i.e. non-voice data transfer); UMTS enables alarger number of subscribers (simultaneous) access tohigh-speed data transfer services. Additionally, UMTS

provides a multifunctionality of services (e.g. mobiletelephony, sms, mms, video telephony, television, fax,

57 UMTS (R'99) networks were not considered as this standard will soonbecome phased out. UMTS (R'06) networks were not presented here as even

slight increases in the network load lead to a reduced environmental impact ascompared with GSM-EDGE.

Table 2Parameters varied in the sensitivity analysis

Networks Parameters Scenarios

0 01 01a 02 02a

GSM-EDGE Subscriber [] 6,189,000 4,087,473 4,087,473 1,986,153 1,986,153UMTS R'04 Subscriber [] 1,051,000 2,101,320 2,101,320 4,202,640 4,202,640

Total download volume per subscriber [Mbit/a] 2256 2256 4512 2256 9024

58 This figure has to be considered carefully as the number of NodeB highlydepends on the transmission technique an operator selects.


ARTICLE IN PRESS


13/20

web-browsing, ftp-services, etc.59) that is difficult or impossible to provide with GSM. Therefore, and inorder to get comprehensive information, the considerationof the absolute impact and the environmental impact per

functional unit is necessary.

6. Conclusions and outlook

Based on the study results, the following conclusions aredrawn and recommendations are made:

a.) A parallel operation of GSM and UMTS networks isenvironmentally adverse (Figs. 6 and 7) and thus thetransition phase between GSM and UMTS technologyshould be kept as short as possible.

b.) It is of urgent importance to lower the operational

energy consumption (HVAC) of the radio networkcomponents, i.e. of the RNS60 (in particular the

NodeB) and of the BSS61 (in particular the BTS). Forinstance, in comparison with public railway transport,the energy demand of mobile phone networksaccounts for almost half of the energy consumed byrailway: the annual operation of one GSM mobile

phone62 (GPRS-compliant) accounts for an energy

consumption of about 530 MJ/(mobile phone year),resulting in a total energy consumption of 916 GWh/year. The Schweizerische Bundesbahnen (Swiss Fed-eral Railways) consume in total about 2000 GWh/year

(P.S., 2005).The reduction of the energy demand of the UMTSradio network components is also of particularimportance in consideration of the fact that UMTSnetworks will be implemented in a GSM-like extentand that NodeB consumes up to six times more energythan BTS.

c.) In order to lower the environmental impact of theproduction phase there is a particular need to thoroughlyanalyse the energy intensive PWBA manufacturing for

potentials to save energy. This could be beneficial inparticular with respect to mobile phones as they contain

large amounts of PWBA (Table 3).Also, a similar simple upgradeability of the UMTSnetworks (3G networks) to next generation networkstandards (e.g. Super 3G, 4G) as it is presently thecase for GSM networks when upgraded withEDGE-technique (ERICSSON, 2003) could help tolower the total environmental impact of thenetworks.

d.) The particular environmental benefit of the EOL phaselies in the fabrication of secondary raw materials, whichcan help to significantly lower the environmental impactof the production phase, in particular with respect to thedepletion of resources. But recycling also has aneconomic aspect. The increasing shortage of, e.g.,

IMPACT2002+

Scenario 0 Scenario 2(a)Scenario 1(a)

Fig. 6. Total impact scores per year relative to the total impact scores of the GSM-EDGE reference network for all four damage categories. For the GSM-EDGEreference network: 7.9413E9 MJ/year non-renewable energy, 1.3776E8 kg CO2-equivalents/year, 93.729 DALY/year and 1.245E8 PDFm

2year/year.

59 Of course, the usefulness of the services is debatable.60 Radio Network Subsystem (in a UMTS network, see Fig. 2).61 B

aseS

tationS

ubsystem (in a GSM network, see Fig. 2).62 Including the energy consumption of the phone and the network.


ARTICLE IN PRESS


14/20

precious and rare metals, such as gold, silver, palladiumand indium results in impressively rising costs making the

production of for example mobile phones ever moreexpansive. Therefore, the treatment of electronic scrapand the recovery of precious and rare metals as well as

energy represent a preferable ecological and economicalternative.

From a methodological point of view the resultsobtained for the environmental effects associated with theEOL phase have to be interpreted carefully. First of all,

present LCA methodology considers overall integratedemissions accumulated assuming a linear model at lowdoses. It does not take into account possible changes indoseresponses at low exposures and natural backgroundconcentrations, for example of metals, are not accountedfor. Secondly, the databases and the impact assessment

method (IMPACT2002+) applied distinguish only betweena very few metal speciations. That can possibly lead to a

biased and blanket characterisation of the environmentalimpact of several metals/metal speciations. In order to

improve the LCA methodology the following futureresearch activities are recommended:

i.) A differentiation of the characterisation factors for metalsin the life cycle impact assessment and speciation of the

metals in the life cycle inventory.ii.) An integration of characterisation factors for flame

retardants.iii.) A realistic approximation of long-term effects of emis-

sions from landfill sites.iv.) The implementation of such study results into the daily

business of telecommunications industry.

With respect to the modelling of large technologicalsystems, such as mobile phone networks, it has been proventhat the modelling of an entire network can be focussed to themodelling of the mobile stations (i.e. mobile phones), the

antenna stations (BTS/NodeB) and the antenna stationcontrollers (BSC/RNC). The core network components (i.e.MSC, SGSN, GGSN) do not need to be modelled in greatdetail as their impact share is comparatively low.

Table 3Comparing PWBA masses contained in GSM network components with components of other ICT a (all figure in the table taken from (Scharnhorst, 2005)

MS BTS BSC MSC TV-set PC

Number of units in Western Europe (estimated for the end of 2005) 3.75E+08 4.17E+05 5210 2605 1.00E+08 2.00E+08Average use time [years] 1.5 7 8 10 10.5 35PWBA [g/unit] 30 31,350 128,000 74,750 500 400PWBA total [t] 11,250 13,060 666 195 50,000 80,000PWBA total [t/year] 7500 1866 83 20 4762 26,66716,000

a Information and Communication Technologies.

Scenario 0

Scenario 1(a) Scenario 2(a)

IMPACT2002+

Fig. 7. Relative impact scores for all four damage categories relative to the impact scores of the GSM-EDGE reference network.


ARTICLE IN PRESS


15/20

Appendix A. Typical architecture of mobile phone networks

Network generation 2G 2.5G 2.5G 3G 3G 3GNetwork standards GSM

(Phase 2)(Phase 2+) (Phase 2+) UMTS

(3GPP R'99)UMTS(3GPP R'04)

UMTS(3GPP R'06)

Data transmission mode(s) CS1 CS/PS2 CS/PS CS/PS CS/PS PS3

Data transmission services GPRS EDGE

Data transfer rates [kbit/s] Voice transmission 9.614.4 9.614.4 9.614.4 12.2 12.2 12.2Data (uplink) 9.6 31.2 192.0 64.0 960.0


16/20

MSC racks (46)6 4 Air conditioner6 8 Cabling6 40 Serving GPRS Support Node (SGSN) 53 23 33SGSN racks7 1 1Cabling ([m], outdoor)7 35 35

Network generation 2G 2.5G 2.5G 3G 3G 3GNetwork standards GSM UMTS UMTS UMTSData transmission

services/Access methodsGPRS EDGE FDD FDD/TDD Fdd/TDD (incl.

HSDPA, USDPA)Network configurations Gateway GPRS Support Node (GGSN) 53 23

GGSN racks8 1 1Cabling ([m], outdoor)8 35 35

1The BTS were assumed to cover three sectors.2In case of UMTS (R'99) the NodeB were assumed to cover three sectors, each comprising one cell (i.e. 1 + 1 + 1). In case of UMTS (R'04) and (R'06) the NodeB wereassumed to cover three sectors, each comprising two cells (i.e. 2 + 2 + 2) Holma, H. and Toskala, A.: WCDMA for UMTS: Radio Access for Third Generation MobileCommunications, John Wiley and Sons Ltd. Chichester. 2004.3Per BTS.4The number of voice channels decreases from 66 per cell to 54 due to information overhead related to Voice overIP (VoIP).5

Per BSC/RNC.6Per MSC.7Per SGSN.8Per GGSN.

Appendix C. Data sources of network components

Network Component Subcomponent Sources

GSM 900 (basic) MS (ERICSSON, 1999, Nokia, 2005a,b)BTS Antenna (Conquadrat, 2003, Doradus, 2003a,b)

Mast (ITF, 2005)Cable (outdoor) (SuperiorCables, 2003)Rack (LucentTechnologies, 2000; SiemensAG, 2000)

Backup batteryBSC Rack (Enderin et al., 2001; ERICSSON, 2002, 2004)MSC Rack (LucentTechnologies, 2001a, 2003a,b,c, 2005, NSI, 2005, SUNmicrosystems, 1999)

GSM 900 (GPRS/EDGE) SGSN Rack (Lucent, 2004a)GGSN Rack (CiscoSystems, 2004, Lucent, 2004b)

UMTS Ue (Anonymous, 2005b, UMTSlink.at, 2004, Xonio, 2004, ZDNet, 2005)NodeB Antenna See in GSM 900 (basic): antenna

Mast See in GSM 900 (basic): mast Cable (outdoor) See in GSM 900 (basic): cable (outdoor)Rack (LucentTechnologies, 2001b, 2004)Back-up battery See in GSM 900 (basic): back-up battery

RNC Rack (Gestner and Persson, 2002)MSC Rack See in GSM 900 (basic): MSCSGSN Rack See in GSM 900 (GPRS/EDGE): SGSNGGSN Rack See in GSM 900 (GPRS/EDGE): GGSN

Appendix D

D.1. Transfer coefficients and fractions for incineration processes

Element Transfer coefficient

Emissionpath

Transferfraction [%]


Emissionpath


Ag_tot 0.01 Ag_air 0.9 Ge_tot 0.01 Ge_air 0.1Ag_soil 0.05 Ge_soil 0.7Ag_wat 0.05 Ge_wat 0.2

Al_tot 0.144 Al_air 0.95 Hg_tot 0.75 Hg_air 0.99Al_soil 0.025 Hg_soil 0.005Al_wat 0.025 Hg_wat 0.005

Appendix B (continued)


ARTICLE IN PRESS


17/20

As_tot 1 As_air 0.99 In_tot 0.01 In_air 0.1As_soil 0.005 In_soil 0.6As_wat 0.005 In_wat 0.3

Au_tot 0.01 Au_air 0.4 Mn_tot 0.1 Mn_air 0.1Au_soil 0.4 Mn_soil 0.6Au_wat 0.2 Mn_wat 0.3

Be_tot 0.01 Be_air 0.9 Na_tot 0.5 Na_air 0.4Be_soil 0.05 Na_soil 0.1Be_wat 0.05 Na_wat 0.5

Bi_tot 0.01 Bi_air 0.1 Ni_tot 0.01 Ni_air 0.1Bi_soil 0.8 Ni_soil 0.6Bi_wat 0.1 Ni_wat 0.3

Br_tot 1 Br_air 0.9 Pb_tot 0.271 Pb_air 0.1Br_soil 0.05 Pb_soil 0.6Br_wat 0.05 Pb_wat 0.3

Cd_tot 1 Cd_air 0.8 Pd_tot 0.01 Pd_air 0.1Cd_soil 0.15 Pd_soil 0.6Cd_wat 0.05 Pd_wat 0.3

Cl_tot 0.75 Cl_air 0.5 Pt_tot 0.01 Pt_air 0.1Cl_soil 0.1 Pt_soil 0.7Cl_wat 0.4 Pt_wat 0.2

Co_tot 0.2 Co_air 0.1 Ru_tot 0.01 Ru_air 0.1Co_soil 0.6 Ru_soil 0.6Co_wat 0.3 Ru_wat 0.3

Cr_tot 0.01 Cr_air 0.2 Sb_tot 0.361 Sb_air 0.2Cr_soil 0.7 Sb_soil 0.6Cr_wat 0.1 Sb_wat 0.2

Cr_VI_tot 0.1 Cr_VI_air 0.3 Se_tot 0.01 Se_air 0.1Cr_VI_soil 0.25 Se_soil 0.7Cr_VI_wat 0.1 Se_wat 0.2

Cu_tot 0.01 Cu_air 0.1 Si_tot 0.2 Si_air 0.05Cu_soil 0.8 Si_soil 0.4Cu_wat 0.1 Si_wat 0.3

Eu_tot 0.01 Eu_air 0.1 Sn_tot 0.01 Sn_air 0.1Eu_soil 0.7 Sn_soil 0.6Eu_wat 0.2 Sn_wat 0.3

Fe_tot 0.041 Fe_air 0.1 Th_tot 0.01 Th_air 0.1Fe_soil 0.6 Th_soil 0.6Fe_wat 0.3 Th_wat 0.3

Ga_tot 0.01 Ga_air 0.2 Zn_tot 0.356 Zn_air 0.1Ga_soil 0.4 Zn_soil 0.7Ga_wat 0.4 Zn_wat 0.2

D.2. Transfer coefficients and fractions for landfill processes


Emissionpath



Emissionpath


Ag_tot 1 Ag_air 0.05 Hg_tot 1 Hg_air 0.22

Ag_soil 0.3 Hg_soil 0.55Ag_wat 0.65 Hg_wat 0.33Al_tot 1 Al_air 0.05 In_tot 1 In_air 0.01

Al_soil 0.3 In_soil 0.65Al_wat 0.25 In_wat 0.34

As_tot 1 As_air 0.1 Mn_tot 1 Mn_air 0.05As_soil 0.3 Mn_soil 0.35As_wat 0.6 Mn_wat 0.6

Au_tot 1 Au_air 0.01 Na_tot 1 Na_air 0.3Au_soil 0.24 Na_soil 0.1Au_wat 0.75 Na_wat 0.6

Be_tot 1 Be_air 0.05 Ni_tot 1 Ni_air 0.01Be_soil 0.4 Ni_soil 0.55Be_wat 0.55 Ni_wat 0.14

Bi_tot 1 Bi_air 0.01 Pb_tot 1 Pb_air 0.05

Bi_soil 0.44 Pb_soil 0.34(continued on next page)

Appendix D.1 (continued)


ARTICLE IN PRESS


18/20

Appendix D.2 (continued)

Bi_wat 0.55 Pb_wat 0.2Br_tot 1 Br_air 0.1 Pd_tot 1 Pd_air 0.04

Br_soil 0.2 Pd_soil 0.36Br_wat 0.7 Pd_wat 0.6

Cd_tot 1 Cd_air 0.05 Pt_tot 1 Pt_air 0.01Cd_soil 0.45 Pt_soil 0.34

Cd_wat 0.35 Pt_wat 0.65Cl_tot 1 Cl_air 0.24 Ru_tot 1 Ru_air 0.05

Cl_soil 0.3 Ru_soil 0.35Cl_wat 0.46 Ru_wat 0.6

Co_tot 1 Co_air 0.01 Sb_tot 1 Sb_air 0.05Co_soil 0.24 Sb_soil 0.75Co_wat 0.75 Sb_wat 0.2

Cr_tot 0.25 Cr_air 0.05 Se_tot 1 Se_air 0.05Cr_soil 0.35 Se_soil 0.65Cr_wat 0.25 Se_wat 0.3

Cr_VI_tot 0.25 Cr_VI_air 0.05 Si_tot 1 Si_air 0.05Cr_VI_soil 0.25 Si_soil 0.85Cr_VI_wat 0.7 Si_wat 0.1

Cu_tot 1 Cu_air 0.05 Sn_tot 1 Sn_air 0.01Cu_soil 0.45 Sn_soil 0.34

Cu_wat 0.35 Sn_wat 0.65Eu_tot 1 Eu_air 0.1 Th_tot 1 Th_air 0.01

Eu_soil 0.3 Th_soil 0.29Eu_wat 0.6 Th_wat 0.7

Fe_tot 1 Fe_air 0.01 Zn_tot 1 Zn_air 0.01Fe_soil 0.14 Zn_soil 0.65Fe_wat 0.85 Zn_wat 0.24

Ga_tot 1 Ga_air 0.05Ga_soil 0.25Ga_wat 0.7

Ge_tot 1 Ge_air 0.01Ge_soil 0.65Ge_wat 0.34

References

AFU (Amt fr Umwelt und Energie). Stromverbrauch Schweizer Handies,Basel; 2004.

Althaus H-J, Classen M, Blaser S, Jungbluth N. Ecoinvent life cycleinventories for metals. Dbendorf: Swiss Centre for Life Cycle Inventories;2004.

Anonymous: High-Speed Downlink Packet Access, http://en.wikipedia.org/wiki/High-Speed_Downlink_Packet_Access ; 2005a.

Anonymous: Sharp V902 UMTS, http://www.factory-2-you.de/shop/datasheet.php?products_id=112648; 2005b.

BAKOM. Amtliche Fernmeldestatistik 2003. Biel: Bundesamt fr Kommunika-tion; 2004a.

BAKOM. Faktenblatt UMTS v2.20. Bern: Bundesamt fr Kommunikation;2004b.

Banet F-J. Anfrage zu UMTS (R'05), Scharnhorst, W. personal communication(e-mail correspondence (2005-05-23)); 2005a.

Banet F-J. Fragen zu UMTS, Scharnhorst, W. personal communication (e-mailcorrespondence (2005-04-25)); 2005b.

Banet F-J, Grtner A, Tessmar G. UMTS Netztechnik, Dienstarchitektur,Evolution. Bonn: Hthig Telekommunikation; 2004.

BDSV: Stahlschrottein grner Rohstoff, Bundesvereinigung DeutscherStahlrecycling-und Entsorgungsunternehmen Dsseldorf; 2001.

Bekkers R, Smits J. Mobile telecommunications: standards, regulation, andapplications. London: Artech House Boston; 1997.

Benkner T, Stepping C. UMTS Universal Mobile TelecommunicationsSystem, J. Schlembach Fachverlag Weil der Stadt; 2002.

Berzelius. QSL Die umweltfreundliche Bleigewinnung. Stolberg: BerzeliusStolberg GmbH; 1993.

BOLIDEN. Boliden world class recycler. Skelleftehamn: BOLIDEN MineralAB; 2002.

BckmannGmbH: Setztisch Technisches Datenblatt, Mnchengladbach.2001a.

BckmannGmbH: Standardkegelsichter (Saugbetrieb) Technisches Daten-blatt, Mnchengladbach; 2001b.

BckmannGmbH: Zick-Zack Windsichteranlagen Technisches Datenblatt,Mnchengladbach; 2001c.

CEC (Comission of the European Parliament). On the restriction of the use ofcertain hazardous substances in electrical and electronic equipment (RoHS).Brussels; 2003a.

CEC (Comission of the European Parliament). On waste electrical and electronicequipment (WEEE). Brussels; 2003b.

CiscoSystems. Cisco 7609 router enhanced 9-slot chassis. San Jose: CiscoSystems; 2004.

Conquadrat (Conquadrat). Produktkatalog 2003 mit Preisliste, Schwanstetten;2003.

Delpho H. Mobilkommunikation 2009. Basel: Prognos AG; 2005.Doradus. 6015 V patch 2.4 Ghz base station antenna. Stella Doradus Ireland

Ltd.; 2003a.Doradus. 12008 V patch 2.4 Ghz base station antenna. Stella Doradus Ireland

Ltd.; 2003b.Duque-Antn, M.: MobilfunknetzeGrundlagen, Dienste und Protokolle,

Vieweg Braunschweig/Wiesbaden; 2002.Eberspcher J, Vogel H-J, Bettstetter C. GSM Global System for Mobile

CommunicationVermittlung, Dienste und Protokolle in digitalenMobilfunknetzen, B.G. Teubner GmbH Stuttgart, Leipzig, Wiesbaden;2001.

ecoinventCentre: ecoinvent data v1.01, Dbendorf, CH; 2003.Ekvall T, Finnveden G. Allocation in ISO 14041a critical review. J Clean

Prod 2001;9:197208.Enderin M, LeCorney D, Lindberg M, Lindquvist T. Axe 810 the evolution

continues. Ericsson Review, vol. 1; 2001.


ARTICLE IN PRESS
http://en.wikipedia.org/wiki/High-Speed_Downlink_Packet_Accesshttp://en.wikipedia.org/wiki/High-Speed_Downlink_Packet_Accesshttp://en.wikipedia.org/wiki/High-Speed_Downlink_Packet_Accesshttp://en.wikipedia.org/wiki/High-Speed_Downlink_Packet_Access


19/20

ERICSSON. Manual Ericsson A1018s. Stockholm: Ericsson Mobile Commu-nication AB; 1999.

ERICSSON (Ericsson Radio Systems AB). BSC/TRC, BSC and TRC 9.1 product overview, Stockholm; 2002.

ERICSSON (ERICSSON AB). EDGE introduction of high-speed data inGSM/GPRS networks, Stockholm; 2003.

ERICSSON. Axe system description. Stockholm: Ericsson Telecom AB; 2004.

ERICSSON. Ericsson demonstrates WCDMA, HSDPA phase 2 during CTIAWireless; 2005.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2): network architecture (GSM 03.02);1996.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2+): Base Station System-Mobile-services Switching Centre (BSS-MSC) interface interface principles(GSM 08.02 version 6.0.0 release 1997), Sophia Antipolis; 1999a.

ETSI (European Telecommunications Standards Institute). Universal MobileTelecommunication System (UMTS): general UMTS architecture, SophiaAntipolis; 1999b.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2+): Base Station Controller-BaseTransceiver Station (BSC-BTS) interfacelayer 3 specifications (GSM

08.58 version 8.4.0 release 1999), Sophia Antipolis; 2000a.ETSI (European Telecommunications Standards Institute). Digital cellular

telecommunications system (phase 2+): GSM/Univeral Mobile Telecom-munication System UMTS/General Packed Radio Service GPRS: serivcedescription stage 1 (3GPP TS 22.060 version 3.5.0 release 1999), SophiaAntipolis; 2000b.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2+): radio transmission and reception(3GPP TS 05.05 version 6.8.0 release 1997), Sophia Antipolis; 2000c.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2+): rate adaption on the Mobile Station-Base Station System (MS-BSS) interface (3GPP TS 04.21 version 8.3.0release 1999), Sophia Antipolis; 2000d.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2+): General Packet Radio Service

(GPRS) overall description of the GPRS radio interface Stage 2 (GSM03.64 version 6.4.0 release 1997), Sophia Antipolis; 2001a.

ETSI (European Telecommunications Standards Institute). Universal MobileTelecommunication System (UMTS): UTRA high speed downlink packetaccess (3GPP TR 25.950 version 4.0.0 release 4), Sophia Antipolis;2001b.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2+): Base Station System-MobileServices Switching Centre (BSS-MSC) interface interface principles(3GPP TS 08.02 version 8.0.1 release 1999), Sophia Antipolis; 2002a.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (phase 2+): general aspects on the BSS-MSCinterface (3GPP TS 08.01 version 8.0.1 release 1999), Sophia Antipolis;2002b.

ETSI (European Telecommunications Standards Institute). Universal Mobile

Telecommunication System (UMTS): network architecture (3GPP TS23.002 version 3.6.0 release 1999), Sophia Antipolis; 2002c.

ETSI (European Telecommunications Standards Institute). Universal MobileTelecommunication System (UMTS): physical layer general descrip-tion (3GPP TS 25.201 version 3.4.0 release 1999), Sophia Antipolis;2002d.

ETSI (European Telecommunications Standards Institute). Universal MobileTelecommunication System (UMTS): UTRAN Iu interface: general aspectsand principles (3GPP TS 25.410 version 3.8.0 release 1999), SophiaAntipolis; 2002e.

ETSI (European Telecommunications Standards Institute): digital cellulartelecommunications system (phase 2+): GSM/EDGE Radio Access Network(GERAN) overall description stage 2 (3GPP TS 43.051 version 5.10.0release 5), Sophia Antipolis; 2003a.

ETSI (European Telecommunications Standards Institute). Digital cellular

telecommunications system (phase 2+): Univeral Mobile Telecommunica-

tion System (UMTS)/General Packet Radio Service (GPRS): serivcedescription stage 2 (3GPP TS 23.060 version 3.16.0 release 1999),Sophia Antipolis; 2003b.

ETSI (European Telecommunications Standards Institute). Universal MobileTelecommunication System (UMTS): network architecture (3GPP TS23.002 version 4.8.0 release 4), Sophia Antipolis; 2003c.

ETSI (European Telecommunications Standards Institute). Universal Mobile

Telecommunication System (UMTS): spacial channel model for MultipleInput Multiple Output (MIMO) simulations (3GPP TR 25.996 version 6.1.0release 6), Sophia Antipolis; 2003d.

ETSI (European Telecommunications Standards Institute). Universal MobileTelecommunication System (UMTS) UTRA High Speed Downlink PacketAccess (HSDPA): overall description stage 2 (3GPP TS 25.308 version6.3.0 release 6), Sophia Antipolis; 2004.

ETSI (European Telecommunications Standards Institute). Digital cellulartelecommunications system (Phase 2+): Universal Mobile Telecommunica-tion System (UMTS) network architecture (3GPP TS 23.002 version6.7.0 release 6), Sophia Antipolis; 2005.

Faist-Emmenegger M, Frischknecht R, Jungblut N. LCA des MobilfunksystemsUMTS. Uster: ESU Services; 2003.

Faist-Emmenegger M, Frischknecht R, Stutz M, Guggisberg M, Witschi R, OttoT. Life cycle assessment of the mobile communication system UMTS. Int J

LCA Online First 2004:112.Farrell F. Integrated Pollution Prevention and Control (IPCC) reference

document on best available techniques in the non ferrous metals industries.Sevilla: Insitute of Prospective Technological Studies; 2000.

Fishbein BK. (INFROM Inc): Waste in the Wireless World: The Challange ofCell Phones, New York; 2002.

Frischknecht R, Faist-Emmenegger M. Ecoinvent Strommix und Stromnetz.Dbendorf: Swiss Centre for Life Cycle Inventories; 2003.

Frischknecht R, Jungbluth N, Althaus H-J, Doka G, Dones R, Heck T, et al.Ecoinvent overview and methodology. Dbendorf: Swiss Centre for LifeCycle Inventories; 2004.

Grtner A. Datentransferraten in UMTS, Scharnhorst, W. personal communi-cation (e-mail correspondence (2005-05-17)).

Gestner B, Persson B. RNC3810 Ericsson's first WCDMA radio networkcontroller. Ericsson Rev 2002;2.

Goosey M, Kellner R. A scoping study end-of-life printed circuit boards.Yorkshire: Department of Trade and Industry; 2002.

Grunewald A, Gustavsson J. End-of-life Treatment of Radio Base StationTransceivers in Europe, Japan and the USA a Life Cycle AssessmentStudy. Stockholm: Industrial Ecology Royal Institute of Technology/Chemical Engineering Linkping Institute of Technology; 1999.

GSMAssociation: GSMA statistics Q2/04, http://www.gsmworld.com/news/statistics/pdf/gsma_stats_q2_04.pdf; 2004.

GSMworld: GSM subscribers connected, http://www.gsmworld.com/index.shtml; 2005.

Guine JB, Gorre M, Heijungs R, Huppes G, Kleijn R, de Koning A, et al.Life Cycle Assessment; An Operational Guide to the ISO Standards; FinalReport, May 2001, Ministry of Housing, Spatial Planning and Environ-ment (VROM) and Centrum voor Milieukunde (CML), RijksuniversiteitDen Haag and Leiden; 2001.

Halonen T, Romero J, Melero J. GSM, GPRS and EDGE performance.Chichester: John Wiley & Sons Ltd.; 2003.

Hausammann E. Power consumption BTS and NodeB, Scharnhorst, W. personalcommunication (e-mail correspondence (2005-02-04)).

HELCOM (HELCOM). Land-based pollution group6h meeting, Vilnius;2002.

Hillebrand F. GSM and UMTS: the creation of global mobile communication.Chichester: John Wiley & Sons, Ltd.; 2002.

Holma H, Toskala A. WCDMA for UMTS: radio access for third generationmobile communications. Chichester: John Wiley & Sons Ltd; 2004.

Hugentobler S. Beschwerlicher Weg beim Bau von Mobilefunkanlagen. NeueZricher Zeitung; 2000. 2000-12-12.

Ihlenfeld J. NOKIA zeigt HSUPA mehr Upstream-Bandbreite fr UMTS,http://www.golem.de; 2005.

IKP and PE (IKP Universitt Stuttgart und PE Europe GmbH). GaBi4: Software

und Datenbank zur Ganzheitlichen Bilanzierung, Stuttgart; 2003.


ARTICLE IN PRESS
http://www.gsmworld.com/news/statistics/pdf/gsma_stats_q2_04.pdfhttp://www.gsmworld.com/news/statistics/pdf/gsma_stats_q2_04.pdfhttp://www.gsmworld.com/index.shtmlhttp://www.gsmworld.com/index.shtmlhttp://www.golem.de/http://www.golem.de/http://www.gsmworld.com/index.shtmlhttp://www.gsmworld.com/index.shtmlhttp://www.gsmworld.com/news/statistics/pdf/gsma_stats_q2_04.pdfhttp://www.gsmworld.com/news/statistics/pdf/gsma_stats_q2_04.pdf


20/20

ISO. ISO 14040: kobilanz Prinzipien und allgemeine Anforderungen.Brussels: International Organisation for Standardisation; 1997.

ISO. ISO 14041: environmental managementlife cycle assessment goaland scope definition and inventory analysis. Brussels: InternationalOrganisation for Standardisation; 1998a.

ISO. ISO 14042: environmental managementlife cycle assessment lifecycle impact assessment. Brussels: International Organisation for Standar-

disation; 1998b.ISO. ISO 14044: environmental managementlife cycle assessment

requirements and guidelines. Brussels: International Organisation forStandardisation; 2005.

ITF: Selbsttragender Antennenmast der Hhe von 36 m des Durchmessers von1016 mm mit zwei Plattformenschwere Konstruktion, http://www.itf.co.yu/stubovi/stub_teski_ger.html ; 2005.

Janssen W. UMTS- T-Mobile beschleunigt HSDPA-Technik auf 7.2 MBit/s,http://www.at-mix.de/news/996.html; 2005.

Jolliet O, Margni M, Charles R, Humbert S, Payet J, Rebitzer G, et al.IMPACT2002+: a new life cycle impact assessment methodology. Int J LCA2003;8(6):32430.

Kincaid LE, Geibig JR. Printed wiring board cleaner technologies substitutesassessment: making holes conductive. Washington: University of Tennessee;1998.

Lehner T. Recovery of precious metals using plastics as reducing agent.Brussels; 2001.

Lucent. Flexent Serving GPRS Support Node (SGSN) for UMTS. Wiltshire:Bell Labs; 2004a.

Lucent. Lucent Gateway GPRS Support Node (GGSN). Wiltshire: Bell Labs;2004b.

LucentTechnologies. FLEXENTGSM macrocell. Wiltshire: Bell Labs; 2000.LucentTechnologies. Flexent mobile switchingcentre for UMTS. Wiltshire: Bell

Labs; 2001a.LucentTechnologies. Flexent OneBTS NodeB. Wiltshire: Bell Labs;

2001b.LucentTechnologies. APX 1000 universal gateway. Wiltshire: Bell Labs; 2003a.LucentTechnologies. APX 8000 universal gateway. Wiltshire: Bell Labs; 2003b.LucentTechnologies. MAX TNT universal gateway. Wiltshire: Bell Labs;

2003c.

LucentTechnologies. The FLEXENT UMTS macrocell base station family.Wiltshire: Bell Labs; 2004.

LucentTechnologies. PacketStar PSAX 4500 multiservice media gateway.Wiltshire: Bell Labs; 2005.

Ludwig CD, Hellweg S, Stucki S. Municipal solid waste management. Berlin:Springer Verlag; 2003.

Malmodin J, Oliv L, Bergmark P. Life cycle assessment of third generation (3G)wireless telecommunication systems at Ericsson, Stockholm; 2001.

Motorola. Material assay of the GSM 900 handy PWB and GSM 900 BTS rackPWB. Taunusstein: Rapid Environmental Assessment Lab; 2005.

Nokia. Nokia HSDPA solution; 2003.Nokia: Nokia 3210, http://www.nokia.ch/german/phones/phone_models/3210/

features.html; 2005a.Nokia: Nokia 6150, http://www.nokia.ch/german/phones/phone_models/6150/

features.html; 2005b.

NSI. Line interface modules for the stinger DSL access concentrator. SimiValley: Network Systems Integration; 2005.

Orange: Schlsselzahlen, http://www.orange.ch/aboutus/corporateaffairs/keyFigures?ts=1113985534112; 2004.

P.S. Wie die SBB am Strom hngen. Neue Zricher Zeitung; 2005. 23.06.2005.Pehrsson U, Hedblom M-O. Sustainability. Stockholm: Ericsson; 2005.Pennington DW, Potting J, Finnveden G, Lindeijer E, Jolliet O, Rydberg T, et al.

Life cycle assessment part 2: current impact assessment practice. Environ Int2004;30:72139.

Pennington DW, Margni M, Ammann C, Jolliet O. Multimedia fate and humanintake modeling: spatial versus nonspatial insights for chemical emissions inWestern Europe. Environ Sci Technol 2005;39:111928.

PhoneContent.com: IDC finds mobile phone market in need of new growthcatalyts, http://www.phonecontent.com/bm/news/se/663.shtml ; 2004.

RANDA-GROUP. Life cycle assessment of a mobile phone. Telefnica

Servicios Mviles S A ; 2000

Rebitzer G, Ekvall T, Frischknecht R, Hunkeler D, Norris G, Rydberg T, et al.Life cycle assessment part 1: framework, goal and scope definition,inventory analysis, and applications. Environ Int 2004; 30:70120.

Rder A, Bauer C, Dones R. Ecoinvent Kohle. Dbendorf: Swiss Centre forLife Cycle Inventories; 2004.

Rudolf R. GPRS basics Die Grundkonzepte des General Packet RadioService. Nrnberg: T.O.P BusinessInteractive; 2003.

Sanders G, Thorens L, Reisky M, Rulik O, Deylitz S. GPRS networks.Chichester: John Wiley & Sons Ltd.; 2003.

Scharnhorst W. Life-Cycle Assessment of Mobile Telephone Networks, withFocus on the Environmental Relevance of the End-of-Life Phase. PhDthesis, EPFL-ENAC-ISTE Lausanne; 2005.

Scharnhorst W, Althaus H-J, Classen M, Jolliet O, Hilty LM. End of lifetreatment of second generation mobile phone networks: strategies to reducethe environmental impact. Env Imp Ass Rev 2005a;25(5):54066.

Scharnhorst W, Althaus H-J, Hilty L, Jolliet O. Environmental assessment ofEnd-of-Life treatment options for an GSM 900 antenna rack. Int J LCAOnline first 2005b:118.

Scharnhorst W, Ludwig C, Wochele J, Schuler A, Hilty LM, Jolliet O. Heavymetal partitioning from electronic scrap during thermal End-of-Lifetreatment. Environ Sci Technol 2005c.

Schnabel P. Kommunikationstechnik-Fibel. Ludwigsburg; 2003.

Schnabel P. Datenbertragung im Mobilfunknetz, http://www.elektronik-kompendium.de, Ludwigsburg; 2004.

Scholz J. Swisscom Mobile: 30000 bis 40000 UMTS-Kunden, http://www.teltarif.ch/arch/2005/kw22/s4059.html, Berlin; 2005.

Schullitz M. (Arthur D. Little Int. GmbH): Key success factors for M-commerce,Vienna; 2001.

SiemensAG (Siemens AG). BS-240/BS-241 base transceiver stations,Mnchen; 2000.

Steele R, Lee C-C, Gould P. GSM, cdmaONE and 3G systems. Chichester: JohnWiley & Sons Ltd; 2001.

Stengele M. Telephone Call: EOL treatment processes at IMMARK,Scharnhorst, W. personal communication (Telephone call 2004-05-11).

Stromtip.de: Stromverbrauch durch Handy-Ladegerte, http://www.stromtip.de/news/2959/archiv/Kaum-Stromverbrauch-durch-Handy-Ladegeraete.htm ;2000.

SUNmicrosystems. Sun's Netra t1400 and Netra t1405 servers. Palo Alto: SunMicrosystems, Inc.; 1999.

Sunrise: sunrise Facts and Figures 2004, http://www.sunrise.net/facts_figures_04.pdf; 2005.

SuperiorCables (Superior Cables GmbH). Glasfaserkabel, Markt Schwaben;2003.

Swisscom (Swisscom). Infrastructure UMTS; 2004.Swisscom: Fakten und Zahlen, http://www.swisscom-mobile.ch/scm/

wir_fakten_und_zahlen-de.aspx; 2005a.Swisscom: Information on mobile phone subscriptions, Scharnhorst, W.

personal communication (Telephone conference (2005-04-18)); 2005b.Tanskanen P, Takala R. Concept of a mobile terminal with active disassembly

mechanism. Helsinki: NOKIA Research Centre; 2001.UMTSlink.at: Motorola V980, http://umtslink.at/cgi-bin/reframer.cgi?/telefone/

3g/motorola_v980.php ; 2004.

Uryu T, Yoshinaga J, Yanagisawa Y. Environmental fate of gallium arsenidesemiconductor disposal. J Ind Ecol 2003;7(2):10312.

VSSV: Information ber Stahlrecycling in der Schweiz, Scharnhorst, W.personal communication (telephone call); 2005.

Weidman E, Lundberg S. Life cycle assessment of ERICSSON third generationsystems. Stockholm: ERICSSON A.B.; 2001.

Weyhe, R.: ACCUREC- Recycling GmbH, Scharnhorst, W. personal commu-nication (internal information (2004-02-16)).

Xonio: Panasonic Z800: UMTS-Handy mit Megapixel-Kamera, http://www.xonio.com/news_12628607.html?tid1=7469 and tid2

2g 3g phone ntworks

Documents